A little help understanding what's going on with the pncconf wizard...

I was getting the following errors mainly on the synced Y, although I also got it on the X during a couple of tests. I've' been playing with this thing most of the evening and am still not happy with the way it is running. I have a configuration that I am more comfortable with than I was before but... (there's always a "but") I am suspicious that the little mini pc that I am running this on may have some issues. I have noticed that after a certain number of times running, stopping, running linuxcnc to test, the performance starts to degrade. A restart of the pc helps but I don't know that I trust this OS install (done way back in May). I am tempted to wipe it and start from scratch - I remember getting it running was a bit of a pain (at the time) and I read elsewhere on this forum that there is a bookworm distro that may be easier to get working that what I went through. On the other hand, it is (mostly) stabile. Maybe I need to dig into stepper tuning further?

I guess I am still unsure about this being a good enough baseline before moving on to Qtplasmac? I guess for starters, I'm going to make a backup of the system so at least I have a starting point if I screw something up badly.

I don't understand why you are getting following errors, especially with such
huge limits (0.5" and 0.05")

I would plot the commanded position, commanded velocity and following error
and perhaps trigger on  errored with halscope to try and diagnose the issue.

That is plot

joint.0.motor-pos-cmd
joint.0.f-error
joint.0.vel-cmd

and trigger on

joint.0.f-errored

I think I should have prefaced the following errors with:

I am getting a "hm2/hm2_7i96s.0: error finishing read" error. I didn't notice this until after reloading Linux on this box. I just spent the last hour reading the back and forth forum.linuxcnc.org/27-driver-boards/4691...error-finishing-read  which I must say is very entertaining but not encouraging.  I've got a few things to confirm and check out on that box and then I'll get back to everyone on this.

OK that's different, that means the the Ethernet latency is bad enough to cause
multiple sequential dropped packets.

Did you run a standard latency test?

What does:

uname -a

report

Whats does

ping 10.10.10.10

report?

Take a gander.

That does not look terrible though you may be better off with a newer
(say 6.4 or 6.5) kernel.

Can you run LinuxCNC and print the results of:

halcmd show param *.tmax

cmorley wrote:

see my advice above.

units in Pncconf are in per minute
units in the INI are per second
960 Inches per minute = 16 inches per second x 60 seconds

Pncconf writes the INI from your selections.
Pncconf defaults to 60 inches a minute (if you are using imperial, something close in metric) which is 1 inch per second.

that's absolutely not a good thing and should be corrected

PCW wrote:

That does not look terrible though you may be better off with a newer
(say 6.4 or 6.5) kernel.

Can you run LinuxCNC and print the results of:

halcmd show param *.tmax

Not until tomorrow - not sure I would know how to install a different kernel...

Well, one day turned into two - here is a screenshot of the hal command window right after jogging around and getting the error(s) (see attached).  If I need to install a different kernel, how would I go about that?

What PC hardware do you have and what Ethernet chip does it use?

will show the network hardware